SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.07347 · CVPR · 2024

Frequency Decoupling for Motion Magnification via Multi-Level Isomorphic Architecture

Fei Wang, Kun Li, Zhun Zhong, Meng Wang, Dan Guo

arXiv · PDF · Open in the Atlas

Code that ran

8 of 14 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (7 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 8 have no contract verdict; 0 failures of our instrument; 6 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2403.07347.json · badge: /api/ran/2403.07347.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
Jiafei127/FD4MM community
Papers with Code archive
8 ran · 6 no recorded run
FunctionStatusWhere it lives
FreqDec Ran
ran on a synthesized input · constructed an object; no result was computed
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("3872816527083689")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, FreqDec: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

class FreqDec(nn.Module):
    """Frequency Decomposition"""

    def __init__(self, channel_num):
        super(FreqDec, self).__init__()
        self.pwconv = nn.Conv2d(channel_num, channel_num * 2, kernel_size=1)
        self.dwconv = nn.Conv2d(channel_num * 2, channel_num * 2, kernel_size=3, stride=1, padding=1, groups=channel_num * 2)
        self.dila1 = nn.Conv2d(channel_num, channel_num, groups=channel_num, kernel_size=3, stride=1, padding=1)
        self.dila2 = nn.Conv2d(channel_num, channel_num, groups=channel_num, kernel_size=3, stride=1, padding=2, dilation=2)
        self.R = nn.GELU()

    def forward(self, x):
        x = self.pwconv(x)
        x = self.dwconv(x)
        x1, x2 = x.chunk(2, dim=1)
        lo = self.dila2(x1)
        hi = self.dila1(x2) - lo
        return (self.R(hi), self.R(lo))

ARGS = {
    "channel_num": 4
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type FreqDec — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a FreqDec output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
FreqMixer Ran
ran on a synthesized input · constructed an object; no result was computed
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("24fd50b60dd3dad4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, FreqMixer: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from einops import rearrange

torch.manual_seed(0)

class FreqDec(nn.Module):
    """Frequency Decomposition"""

    def __init__(self, channel_num):
        super(FreqDec, self).__init__()
        self.pwconv = nn.Conv2d(channel_num, channel_num * 2, kernel_size=1)
        self.dwconv = nn.Conv2d(channel_num * 2, channel_num * 2, kernel_size=3, stride=1, padding=1, groups=channel_num * 2)
        self.dila1 = nn.Conv2d(channel_num, channel_num, groups=channel_num, kernel_size=3, stride=1, padding=1)
        self.dila2 = nn.Conv2d(channel_num, channel_num, groups=channel_num, kernel_size=3, stride=1, padding=2, dilation=2)
        self.R = nn.GELU()

    def forward(self, x):
        x = self.pwconv(x)
        x = self.dwconv(x)
        x1, x2 = x.chunk(2, dim=1)
        lo = self.dila2(x1)
        hi = self.dila1(x2) - lo
        return (self.R(hi), self.R(lo))

class FreqMixer(nn.Module):

    def __init__(self, dim, num_heads, bias):
        super(FreqMixer, self).__init__()
        self.dim = dim
        self.num_heads = num_heads
        self.temperature = nn.Parameter(torch.ones(num_heads, 1, 1))
        self.fd = FreqDec(self.dim)
        self.project_out = nn.Conv2d(dim, dim, kernel_size=1, bias=bias)
        self.act = nn.ReLU()

    def forward(self, x):
        b, c, h, w = x.shape
        hi, lo = self.fd(x)
        low_q = rearrange(lo, 'b (head c) h w -> b head c (h w)', head=self.num_heads)
        k = rearrange(hi, 'b (head c) h w -> b head c (h w)', head=self.num_heads)
        v = rearrange(hi, 'b (head c) h w -> b head c (h w)', head=self.num_heads)
        low_q = torch.nn.functional.normalize(low_q, dim=-1)
        k = torch.nn.functional.normalize(k, dim=-1)
        attn = low_q @ k.transpose(-2, -1) * self.temperature
        attn = self.act(attn)
        out = attn @ v
        out = rearrange(out, 'b head c (h w) -> b (head c) h w', head=self.num_heads, h=h, w=w)
        out = self.project_out(out)
        return out

ARGS = {
    "dim": 4,
    "num_heads": 2,
    "bias": True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
einops, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type FreqMixer — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a FreqMixer output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
HighpassMixer Ran
ran on a synthesized input
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("7ba83054b670e294")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np
from einops import rearrange

INIT_ARGS = {
    'dim': 4,
    'num_heads': 2,
    'ffn_expansion_factor': 2
}

# Create input tensor: batch=2, channels=4, height=3, width=4
input_tensor = torch.randn(2, 4, 3, 4)

ARGS = {
    'x': input_tensor
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
einops, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
Mlp Ran
ran on a synthesized input · constructed an object; no result was computed
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("97cc636e2d3de32c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, Mlp: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Construct arguments for Mlp.__init__
ARGS = {
    'dim': 64,
    'ffn_expansion_factor': 2,
    'act_layer': nn.GELU,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Mlp — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Mlp output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
OverlapPatchEmbed Ran
ran on a synthesized input · constructed an object; no result was computed
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("4c8394b0b6dd4d83")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, OverlapPatchEmbed: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# OverlapPatchEmbed.__init__ takes: in_c, embed_dim, bias
# OverlapPatchEmbed.forward takes: x (a tensor)

# Create a small input tensor: (batch=2, channels=3, height=8, width=8)
x = torch.randn(2, 3, 8, 8)

ARGS = {
    "in_c": 3,
    "embed_dim": 4,
    "bias": False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type OverlapPatchEmbed — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a OverlapPatchEmbed output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
Sparse_Highpass_Filter Ran
ran on a synthesized input · constructed an object; no result was computed
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("0b72fc553533515e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, Sparse_Highpass_Filter: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    "dim": 4,
    "num_heads": 2,
    "bias": True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
einops, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Sparse_Highpass_Filter — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Sparse_Highpass_Filter output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
Sparse_Lowpass_Filter Ran
ran on a synthesized input · constructed an object; no result was computed
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("5d4d8dd08a427ba4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, Sparse_Lowpass_Filter: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The function is Sparse_Lowpass_Filter.__init__ with parameters: self, dim, num_heads, bias
# We need to provide values for dim, num_heads, and bias

ARGS = {
    'dim': 4,
    'num_heads': 2,
    'bias': True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
einops, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Sparse_Lowpass_Filter — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Sparse_Lowpass_Filter output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
Upsample Ran
ran on a synthesized input · constructed an object; no result was computed
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("326a20276e854360")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, Upsample: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The Upsample module takes a single parameter: dim
# dim is used as the input/output channel dimension for Conv2d
# We'll use a small dimension value

dim = 4

ARGS = {
    "dim": dim
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Upsample — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Upsample output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
AFreqMixer Not yet run
unverified (harvested, no recorded run)
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("009cebf0b808d5b7")
How it ran

no invocation recorded.

Freq_Pyramid Not yet run
unverified (harvested, no recorded run)
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("94eca4cce92e1345")
How it ran

no invocation recorded.

LowpassMixer Not yet run
unverified (harvested, no recorded run)
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("1428acfae0d9c3bd")
How it ran

no invocation recorded.

MagNet Not yet run
unverified (harvested, no recorded run)
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("b8c2127e218da170")
How it ran

no invocation recorded.

Manipulator Not yet run
unverified (harvested, no recorded run)
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("ca2b178bfdb9c0fb")
How it ran

no invocation recorded.

Pyramid_recons Not yet run
unverified (harvested, no recorded run)
Jiafei127/FD4MM/magnet_FD4MM.py
code served (permissive licence) · get_code("e850aff262afad31")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 14 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Learning that separates shape and texture or Multi-domain Learning from phase fluctuations. Inspired by the frequency spectrum, we observe that the low-frequency components with stable energy always possess spatial structure and less noise, making them suitable for modeling the subtle motion field. To this end, we present FD4MM, a new paradigm of Frequency Decoupling for Motion Magnification with a Multi-level Isomorphic Architecture to capture multi-level high-frequency details and a stable low-frequency structure (motion field) in video space. Since high-frequency details and subtle motions are susceptible to information degradation due to their inherent subtlety and unavoidable external interference from noise, we carefully design Sparse High/Low-pass Filters to enhance the integrity of details and motion structures, and a Sparse Frequency Mixer to promote seamless recoupling. Besides, we innovatively design a contrastive regularization for this task to strengthen the model's ability to discriminate irrelevant features, reducing undesired motion magnification. Extensive experiments on both Real-world and Synthetic Datasets show that our FD4MM outperforms SOTA methods. Meanwhile, FD4MM reduces FLOPs by 1.63× and boosts inference speed by 1.68× than the latest method. Our code is available at https://github.com/Jiafei127/FD4MM .

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2403.07347")
get_code_for_paper("2403.07347")
have("2403.07347")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2403.07347.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2403.07347.svg)](https://syntology.ai/paper/2403.07347)

Connect an agent — have() is free.